Neisseria meningitidis NalP cleaves humancomplement C3, facilitating degradation of C3band survival in human serumElena Del Tordelloa,1, Irene Vaccaa, Sanjay Ramb, Rino Rappuolia,2, and Davide Serrutoa,2

aResearch Center, Novartis Vaccines and Diagnostics, 53100 Siena, Italy; and bDivision of Infectious Diseases and Immunology, University of MassachusettsMedical School, Worcester, MA 01605

Contributed by Rino Rappuoli, November 28, 2013 (sent for review July 15, 2013)

The complement system is a crucial component of the innate im-mune response against invading bacterial pathogens. The humanpathogen Neisseria meningitidis (Nm) is known to possess severalmechanisms to evade the complement system, including bindingto complement inhibitors. In this study, we describe an additionalmechanism used by Nm to evade the complement system and sur-vive in human blood. Using an isogenic NalP deletion mutant andNalP complementing strains, we show that the autotransporter pro-tease NalP cleaves C3, the central component of the complementcascade. The cleavage occurs 4 aa upstream from the natural C3cleavage site and produces shorter C3a-like and longer C3b-like frag-ments. The C3b-like fragment is degraded in the presence of thecomplement regulators (factors H and I), and this degradationresults in lower deposition of C3b on the bacterial surface. We con-clude that NalP is an important factor to increase the survival of Nmin human serum.

serum resistance | immune evasion

The Gram-negative capsulated bacterium Neisseria meningitidis(Nm) is one of the major causes of meningitis and septicemia

worldwide (1). During the course of infection, the bacterium mustadapt to different host niches, a crucial factor for its survivaland dissemination (2). Evasion of the complement system iscritical for Nm to cause invasive disease. The observation thatpeople deficient in various complement components are highlypredisposed to invasive meningococcal disease provides epide-miologic evidence for the role of complement in host defenseagainst this bacterial infection (3, 4). Complement activationcan occur through the classical pathway (CP), the lectin pathway,or the alternative pathway (AP). Activation of the CP and lectinpathway results in cleavage of C4; deposited C4b can bind to C2,and cleavage of the latter generates C4b,2a, which functionsas a C3 convertase. Cleavage of C3 by convertases is a criticalevent in complement activation, because it leads to release ofthe anaphylatoxin C3a and deposition of C3b on the bacterialsurface (5). Strict regulation of the complement system is nec-essary to avoid inappropriate activation and host cell damage.As an example, C3b in the fluid phase is cleaved to iC3b byfactor I (fI) in the presence of the cofactor factor H (fH). Thiscleavage inactivates C3b and prevents C3b from forming AP C3convertase or AP/CP C5 convertase enzymes (6). The ability toescape to the complement system is a key determinant in thevirulence of pathogens. Bacteria can escape recognition by thecomplement system through the actions of cell surface structuresor secreted proteins (7–9). Nm has evolved several redundantmechanisms to evade the host innate responses at sites of colo-nization and during systemic growth, including expression of fHbp(10, 11) and NspA (12, 13), that facilitate binding of fH, lip-ooligosaccharide (LOS) sialic acid, which inhibits complementdeposition (12), and Opc protein, which binds to vitronectin (2).In a recent functional genomic study, we identified factors in-volved in Nm survival in human blood (14). One of these factors isNalP, which is an autotransporter with subtilisin-like serine

protease activity that is involved in autoproteolytic processing,resulting in secretion of the NalP passenger domain into thebacterial supernatant (15, 16). The expression of NalP is phase-variable because of slipped-strand mispairing of a polycytidinetract in the coding sequence (17). NalP contains a lipobox at theC-terminal end of the signal sequence; the lipid moiety permitsanchorage in the outer membrane (16). Lipidated NalP proteins areonly temporarily retained at the cell surface; however, the lipidmoiety retards the release of NalP passenger domain from thebacterial surface and allows the partial or total cleavage of surfaceprotein targets on bacterial surface (18), including IgA protease Iga,adhesion and penetration protein App (16), autotransporter serineprotease AusI (19), Neisserial heparin binding antigen NHBA (20),and lactoferrin binding protein LbpB (21). This activity modulatesthe expression of meningococcal proteins at bacterial surface andalso has recently been implicated in the formation and regulation ofNm biofilm (22). Although several bacterial targets of NalP pro-tease have been well-characterized, host targets of NalP have notbeen identified thus far. In this work, we elucidate the molecularbasis of Nm complement resistance mediated by NalP.

ResultsNalP Protease, and Not Its Protein Targets, Is Essential for Survival inHuman Serum. In a previous study of Nm transcriptome analysis inhuman blood, we showed that the nalP gene was up-regulatedduring incubation in human blood and that survival of a nalPdeletion mutant (named MCΔnalP) in human blood was im-paired (14). To further characterize the NalP phenotype, we

Significance

The complement system is a crucial component of the innateimmune response in humans. In this study, we report thecharacterization of an autotransporter protease of Neisseriameningitidis named NalP. We show that NalP is able to cleavethe α-chain of the human complement factor C3 in a species-specific manner. As a consequence, the deposition of C3b onthe bacterial surface is reduced and, in human serum, the NalP-generated C3b fragment is further degraded by host factors.Our results significantly increase the knowledge of how N.meningitidis can survive and multiply in human serum, evadingthe innate immune system.

Author contributions: E.D.T., S.R., R.R., and D.S. designed research; E.D.T. and I.V. per-formed research; S.R. contributed new reagents/analytic tools; E.D.T., I.V., S.R., R.R., andD.S. analyzed data; and E.D.T., S.R., R.R., and D.S. wrote the paper.

Conflict of interest statement: R.R. and D.S. are full-time employees of Novartis Vaccinesand Diagnostics.

Freely available online through the PNAS open access option.1Present address: Microbiology and Immunobiology Department, Harvard MedicalSchool, Boston, MA 01255.

2To whom correspondence may be addressed. E-mail: rino.rappuoli@novartis.com ordavide.serruto@novartis.com.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1321556111/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1321556111 PNAS | January 7, 2014 | vol. 111 | no. 1 | 427–432

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

May

23,

202

0

complemented the strain MCΔnalP with the shuttle vector pFP-12expressing nalP under the control of a constitutive promoter togenerate the complementing strain MCΔ–CnalP. The expressionof NalP in the three different isogenic strains was evaluated byWestern blot analysis. As shown in Fig. 1A, the complementingstrain expressed NalP in its active form. Three different specificfragments were detected in whole-cell protein lysates: the full-length form (∼110 kDa), a ∼70-kDa fragment corresponding tothe passenger domain that also contains the catalytic domain, andthe ∼30-kDa translocator domain fragment (15, 16). The presenceof only the passenger domain but not intact NalP or the trans-locator domain in the supernatants prepared from the NalP-expressing strains was confirmed by Western blot analysis (Fig.1B). The complementing strain MCΔ−CnalP expresses higheramounts of NalP compared with the WT strain, likely because ofthe episomal complementation system used. The WT, mutant, andcomplementing strains were tested for survival in human serum.As shown in Fig. 1C, the strain lacking NalP was killed in human

serum, and survival was restored in the complementing strain,confirming a role for NalP in the survival of Nm in human serum.Heat-inactivated serum (all complement pathways inactivated)was used as control and showed similar survival for all strains,confirming a role for complement in killing of the NalP mutant(Fig. 1D). The survival of the three isogenic Nm strains in humanwhole blood paralleled with the phenotype observed in humanserum (Fig. S1). NalP is a surface-exposed autotransporter withserine protease activity able to cleave different surface-exposedproteins, including Iga, NHBA, AusI, App, and LbpB (15, 16, 20,21). To determine whether NalP contributed to Nm survival inblood through an indirect effect by cleaving one or more of theknown target proteins, deletion mutants of each NalP target weregenerated individually in strain MC58. Because nalP is prone tophase variation (15, 16), each mutant strain was tested for NalPexpression. Western blot confirmed that all of the mutant strainsexpress comparable levels of NalP (Fig. S2A). The WT MC58 andmutant strains were then incubated in human whole blood for 2 h.Samples were obtained at various time points to assess the numberof viable colonies (colony forming unit, cfu). As shown in Fig. S2B,bacterial survival for each mutant strain was similar to the MC58WT strain. This result suggests that the phenotype observed inhuman blood for the nalP deletion mutant strain is not caused bythe cleavage of one of the known NalP targets.

NalP Specifically Cleaves Human C3. Looking for additional NalPtargets, we explored the possibility that NalP could inactivatecomponents of the complement system. The WT MC58 strainand its MCΔnalP deletion mutant and MCΔ−CnalP comple-menting strain were examined for their ability to cleave com-plement proteins C1q, C2, C3, and C4, which are involved inactivation of the complement cascade. Nm strains were in-cubated with each of the purified complement molecules for16 h at 37 °C, and the effect of NalP on the complement factorwas determined by Western blot analysis. C1q, C2, and C4 werenot cleaved by any of the strains (Fig. S3). In contrast, incubationof WT MC58 with human C3 generated an ∼100-kDa α-chainfragment that was absent in the reaction containing the MCΔnalPdeletion mutant strain (Fig. 2A); the cleavage was more evidentwith the complementing strain MCΔ−CnalP, which expressed thehighest amounts of NalP. To evaluate whether the passengerdomain of NalP, released in the bacterial supernatant after itsautoproteolytic processing, had similar C3 cleaving activity, con-centrated supernatants from the WT, deletion mutant, and com-plementing MC58 strains were assayed for human C3 cleavage.The passenger domain alone that contains the catalytic activity ofNalP and is released in the supernatant of WT and MCΔ−CnalPcomplementing strains was able to cleave human C3 (Fig. 2B). Weused rabbit polyclonal α-C3/C3a antibody to immunoprecipitateC3 and its fragments from the reaction mixture, and we observedthat a fragment of ∼10 kDa was also generated by bacterialsupernatants from WT and NalP complementing strains. Thisfragment comigrated with the natural C3a (Fig. 2C). To map thecleavage site, the ∼100-kDa C3 fragment generated by NalP wassubjected to N-terminal sequence analysis. The sequence obtained(745-GLARSNLDED-754) showed that NalP cleaved the C3α-chain 4 aa N-terminal to the physiologic cleavage site of C3convertase, which cleaves C3 between Arg748 and Ser749 (Fig.2D). Hence, NalP generated a C3a-like molecule that is 4 aashorter than physiologic C3a and a C3b-like fragment that is 4 aalonger than the natural C3b. Despite several attempts, we wereunable to express in Nm a NalP protein that contained a loss offunction mutation in the catalytic triad. Hence, to confirm thespecificity of the cleavage, the C3 cleavage assay was conducted inthe presence of increasing concentrations of the serine proteaseinhibitor PMSF. We observed a dose-dependent inhibition of C3cleavage, and cleavage was completely abrogated at 5 mM PMSF(Fig. 2E). To further confirm NalP-specific C3 cleavage, we clonedand expressed the WT and mutated nalP genes in Escherichia colistrain BL21-DE3 using the expression vector pET-21b to generatestrain Ec-NalP and Ec-NalPS426A, respectively. Western blot

Fig. 1. Complementation of NalP restores survival in human serum. (A)Western blot analysis showed the expression of NalP in Nm WT and NalPcomplementing strain but not NalP deletion mutant. Three forms of NalP,corresponding to the full-length protein (100 kDa), the released passengerdomain (70 kDa), and the translocator domain (30 kDa), are indicated byarrows. *Not specific bands. A schematic representation of NalP protein withdifferent domains is also shown. The catalytic site of the passenger domain isindicated by a light gray box, and the catalytic triad is indicated by verticalblack lines. (B) The presence of only the passenger domain but not intactNalP or the translocator domain in the supernatant was confirmed byWestern blot analysis. (C) Nm WT, NalP deletion mutant, and complement-ing strains were tested for survival in human serum at 40% over a timecourse of 120 min using serum from two different blood donors. The graphsrepresent the survival in human serum, whereas Insets represent the growthcontrol in GC-rich medium for the same time course of 120 min. Error barsare indicated by vertical lines. (D) Complement is required for killing of theNalP mutant. All strains survived in heat-inactivated human serum (com-plement activity abrogated by heat treatment at 56 °C for 30 min).

428 | www.pnas.org/cgi/doi/10.1073/pnas.1321556111 Del Tordello et al.

Dow

nloa

ded

by g

uest

on

May

23,

202

0

analysis (Fig. S4A) showed that NalP was expressed and localizedin the outer membrane protein fraction in strains Ec-NalP and Ec-NalPS426A; mutation of the serine at position 426 to alanine(S426A) abolished autoprocessing and secretion of the passengerdomain in the supernatant of strain Ec-NalPS426A. Bacterialpellets and concentrated supernatants from cultures of E. colistrains Ec-NalP, Ec-NalPS426A, and Ec-pET (vector control)were tested for their ability to cleave human C3. The C3 α-chainwas cleaved only in the presence of pellet and supernatant fractionprepared from strain Ec-NalP that expressed WTNalP but not thestrain that expressed NalP with the S426A mutation, therebyconfirming specificity of the cleavage mediated by NalP (Fig.S4B). The interaction of Nm with several human proteins has beenshown to be species-specific (13, 23, 24). To investigate whetherthe cleavage of C3 mediated by NalP was restricted to the humanprotein, a cleavage assay was performed using C3 purified fromrabbit and mouse sera. Culture supernatants from Nm strainsMC58, MCΔnalP, and MCΔ−CnalP were incubated with rabbit ormouse C3, and the mixtures were then analyzed by Western blotusing anti-rabbit C3 or anti-mouse C3, respectively. NalP failed tocleave both rabbit and mouse C3, suggesting that NalP cleaves C3in a human-specific manner (Fig. 2 F and G).

NalP Decreases C3b Deposition on the Bacterial Surface. Havingshown that NalP is able to cleave human C3, we hypothesized

that NalP-mediated C3 cleavage would inhibit complementactivation and C3b fragment deposition on the bacterial surface.As shown in Fig. 3A, fewer C3b fragments were deposited onMC58 and MCΔ−CnalP strains compared with the MCΔnalPdeletion mutant strain, which was measured by flow cytometry.Thus, NalP expressed on the bacterial surface is able to cleavehuman C3 and limit deposition of C3b on the bacterial surface.

C3b-Like Molecule Generated by NalP Is Degraded by Serum HostFactors. One possibility to explain the lower deposition of C3bon the bacterial surface observed above is that the NalP-generatedC3b-like fragment is susceptible to degradation by complementregulators. One known mechanism of regulation involves pro-teolysis of C3b by fI in conjunction with its cofactor fH, whichbinds C3b and increases the affinity for fI (6). To test this hy-pothesis, human C3 was first incubated with concentrated super-natant from culture of MCΔ−CnalP to generate the C3b-likecleavage fragment. This mixture was then incubated with humanfH and fI either combined or separately (Fig. 3 B and D, re-spectively) or 2% (vol/vol) human C3-depleted serum that servedas a source of fH and fI (Fig. 3C) for periods of time rangingfrom 0 to 1 h. fH and fI degraded only the C3b-like moleculegenerated by NalP (Fig. 3C, open arrow) but had no effect onintact C3. These results showed specific degradation of only theC3b-like molecule generated by NalP by fH and fI.

Fig. 2. NalP specifically cleaves human C3. (A) Nm WT, NalP deletion mutant, and complementing strain bacterial pellets were incubated with human C3 at37 °C for 16 h. Western blot analysis showed that NalP-expressing strains were only able to cleave C3 α-chain and not C3 β-chain (black arrows), releasinga C3b-like α′ of ∼ 100 kDa (open arrow). (B) The supernatants were tested for their ability to cleave human C3 after 3 or 16 h of incubation. Only bacterialsupernatants containing NalP passenger domain (i.e., WT and complementing strains) were able to cleave human C3 α-chain. Cleavage was evident even after3 h of incubation. (C) A molecule with a size similar to C3a is released from C3 after cleavage by NalP. The reaction mixture was immunoprecipitated usingrabbit polyclonal α-C3/C3a antibody. The Western blot analysis with anti-human C3a showed a C3a-like molecule of ∼ 10 kDa generated only in reactions thatcontained supernatants (SNs) from the WT and complementing strains that possessed the passenger domain of NalP. (D) The C3b-like fragment, generated byNalP, was subjected to N-terminal sequencing. The results revealed that NalP cleaved human C3 between Lys744 and Gly745 (4 aa N-terminal to the C3convertase cleavage site), thereby generating a shorter C3a-like molecule and a longer C3b-like molecule compared with the corresponding physiologicalcounterparts. (E) The C3 cleavage assay using bacterial supernatant obtained from the complementing strain was conducted in the presence of increasingconcentrations of the serine protease inhibitor PMSF ranging from 0.1 to 5 mM. Western blot analysis of samples showed a dose-dependent inhibition ofhuman C3 cleavage by NalP, and the cleavage was completely abrogated at 5 mM PMSF. (F and G) Cleavage of C3 by NalP is human-specific. NalP frombacterial supernatants was tested for its ability to cleave C3 isolated from (F) rabbit and (G) mouse serum by Western blot analysis, and no cleavage wasobserved. In the case of mouse C3, the antibody used does not detect the C3 β-chain.

Del Tordello et al. PNAS | January 7, 2014 | vol. 111 | no. 1 | 429

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

May

23,

202

0

DiscussionNalP is a surface-exposed autotransporter of Nm with serineprotease activity that has been shown to cleave several proteinson the outer membrane of Nm. In this work, we have shown thatNalP also cleaves a host protein and that this cleavage representsa major mechanism for survival in human serum. We provideevidence that NalP is able to cleave human C3 in a human-specific manner. The cleavage occurs 4 aa upstream from thenatural cleavage site and produces a shorter C3a-like and longerC3b-like fragments. The C3b-like fragment generated by NalP issusceptible to degradation by the complement inhibitors, fH andfI. This degradation results in reduced deposition of C3b on thesurface of Nm. We have not investigated whether the C3a-likefragment generated by NalP is active; however, it lacks theC-terminal arginine that is known to be essential for its activity.Therefore, we believe that the NalP cleavage of human C3generates a C3a-like fragment that does not possess the ana-phylatoxin activity of C3a generated physiologically by C3 con-vertases. It is likely that both the decreased C3b deposition andthe absence of the anaphylatoxin activity contribute to Nm im-mune evasion and survival in serum and blood.

Species-Specific Cleavage of Complement Factors. Proteolytic cleav-age of complement components by bacterial proteases is a mech-anism used by many human pathogens. Examples are streptococcalcysteine protease SpeB, which degrades C3 to inhibit bacterialclearance (25), the streptococcal cell-associated peptidase ScpA,which cleaves C5a to inhibit neutrophil chemotaxis (26), GelE ofEnterococcus faecalis (27) and aureolysin of Staphylococcus aureus,which cleave C3 (28), EspP of enterohemorrhagic E. coli, whichcleaves C3 and C5 to impair complement activation (29), and PgtEof Salmonella enterica, which proteolytically cleaves C3b, C4b, andC5 to enhance bacterial resistance to human serum (30).NalP is the only known Nm protein that cleaves C3 and pro-

motes bacterial survival, because a nalP deletion mutant was notable to cleave C3 or survive in human serum or whole blood.However, the phenotype was restored when nalP was com-plemented in trans through episomal expression. Although ourattempts to generate a complementing strain expressing NalP thatwas mutated in the catalytic triad were not successful, the speci-ficity of the catalytic domain of NalP in the C3 cleavage was con-firmed by expressing WT NalP and a mutant NalP that lackedcatalytic activity (S426A) in a nonpathogenicE. coli strain.Wealsoshowed that NalP could cleave human C3 but not rabbit or mouseC3, suggesting that the cleavage is species-specific. Although thefull amino acid sequence of rabbit C3 is not available, alignment ofseveral available nonhuman C3 amino acid sequences, includingmouse C3, shows some diversity in the region surrounding theNalPcleavage site that can explain the species specificity of the cleavage(Fig. S5). The species specificity of NalP is consistent with the otherNm proteins that interact specifically with human factors, whichinclude fHbp, NspA, opacity proteins, and transferring bindingprotein (13, 24, 31, 32). The specificity of interactions with humanfactors may explain why Nm is strictly a human pathogen and whyantibody-mediated killing of Nm in vitro is more efficient whencomplement is derived from rabbits, rats, or mice, with fH and C3that are not recognized by fHbp, NspA, or NalP, respectively.

Molecular Mechanisms for Survival in Serum. It is likely that C3 thatis cleaved by NalP—and particularly, the C3 molecules cleavedin solution by the released passenger domain—does not formcovalent bonds with the bacteria, because the thiolate and acy-limidazole intermediates that are formed on exposure of internalthioester bond in C3 after release of C3a (or C3a-like) are highlylabile with a half-life <100 μs (33) and thus, may undergo hy-drolysis by reacting with a water molecule before encounteringelectron-donating –OH residues on the bacterial surface. Similarto the ability of fH and fI to rapidly cleave C3b that is generatedby C3 convertases (6), the C3b-like molecule generated by NalPis also rapidly degraded by fH and fI. Thus, activation of C3 inthe fluid phase followed by rapid degradation of the C3b-like

molecule by fI and fH (either in solution or bound to the bac-terial surface through ligands, such as fHbp and NspA) couldconstitute a mechanism of complement escape. The C3a mole-cule produced by C3 convertase is one of the anaphylatoxins ofthe complement system and has diverse effects ranging from therelease of histamine from human mast cells to chemotaxis ofbasophils and eosinophils to activation of neutrophils (34–37).The activity of C3a in the bloodstream and tissues is tightlycontrolled by carboxypeptidases, which rapidly remove the C-terminal Arg residue (38, 39). The resulting molecule, calledC3a-desArg, is devoid of proinflammatory activity (40). We,therefore, believe that the C3a-like molecule generated by NalP,which lacks the C-terminal 4 aa, would also lack the biologicalfunctions observed by intact C3a. Similar to NalP, S. aureusaureolysin also cleaves C3 at a single site. Unlike NalP, aur-eolysin cleaves C3 2 aa C-terminal to the C3 convertase cleavagesite to generate functionally active C3b and C3a (28). In contrastto NalP, the streptococcal cysteine protease SpeB fully degradesC3 as a result of cleavage at multiple sites (25).

NalP Expression in Neisseria Species. Based on a bioinformaticsanalysis conducted with the genomic sequences available inthe databases, Nm is the only species of the genus Neisseria thatpossesses a functional nalP (Table S1). nalP is not present inany commensal species but is present in the human pathogenN. gonorrheae. However, none of the N. gonorrheae genomesequences analyzed (17 strains) seem to encode full-length NalPproteins because of the presence of premature stop codons(Table S1). This analysis is in accordance with a previous reportthat nalP in N. gonorrheae FA1090 is a pseudogene because ofthe presence of numerous termination codons (15). In the caseof Nm, the nalP gene was present in all but one of the strains

Fig. 3. Functional characterization of human C3 cleavage by NalP. (A) NmWT, NalP deletion mutant, and the complementing strains were incubatedwith 50% human serum for 60 min. C3b deposition was determined usinggoat polyclonal α-human C3 antibody and analyzed by FACS. The meanfluorescence intensity for the entire bacterial population is shown (text colorcorresponds to the color of the histogram). After 16 h of incubation of hu-man C3 with bacterial supernatant from NalP complementing strain, (B)human fH and fI or (C) human C3-depleted serum (that contains both fH andfI) at 2% (vol/vol) was added. Samples were obtained at the different timepoints indicated, and degradation of the C3b-like fragment generated byNalP (open arrow) was analyzed by Western blotting. (D) Control reactionsincluded samples incubated with fH or fI separately for 1 h.

430 | www.pnas.org/cgi/doi/10.1073/pnas.1321556111 Del Tordello et al.

Dow

nloa

ded

by g

uest

on

May

23,

202

0

analyzed (Table S1). About one-half (13 of 27) of the sequencesanalyzed (listed in Table S1) possess a frameshift in the poly-cytidine region that results in phase variation. It is conceivablethat Nm needs to switch the expression of the nalP gene on or offin different niches of its human host. In the context of adaptationand proliferation in human blood, strains that are phase-offcould revert to phase-on phenotype to express nalP and facilitatethe survival in human blood. We propose that Nm strains thatshow high rates of nalP gene phase variation may possess anadvantage when causing invasive disease in humans.In a recent molecular epidemiology study of 647 Nm isolates,

nalP was not identified in ∼11% of the strains because of deletionevents in the nalP locus (41). The infrequency of the nalP deletionsuggests that it is disadvantageous for the bacteria and that, in thenalP-negative strain, the function could be compensated by otherfactors or othermechanisms. This hypothesis recalls the redundancyin fH binding by Nm, where strains lacking fHbp can use NspA tobind fH (13, 14) to avoid complement activation on bacterial surface.

Role of the NalP-Secreted Passenger Domain and Full-Length Molecule.The full-length NalP protein is exposed on the meningococcalsurface, whereas the passenger domain is released in the super-natant (15, 16). Roussel-Jazédé et al. (18) reported that the lipidmoiety of NalP enables it to be retained at the cell surface fora duration sufficient to cleave its targets on bacterial surface.However, none of the NalP targets seem to contribute to serumresistance of Nm. The majority of NalP on the surface is even-tually autoprocessed, and the passenger domain with its catalyticdomain is released into the supernatant (16, 42). We have shownthat the full-length protein on bacterial surface as well as the re-leased passenger domain both cleave human C3. Inhibition ofcomplement activation by NalP occurs by cleavage of C3, whichgenerates a functionally inactive C3a-like molecule, whereas thecorresponding C3b-like fragment is inactivated rapidly by fH andfI (Fig. 4). For NalP to function as a complement inhibitor, it iscrucial that the passenger domain is secreted in the extracellularenvironment, and therefore, it can cleave C3 distal to the bacterialsurface to generate C3b that is incapable of forming covalentbonds with bacterial targets to generate C3 convertases as de-scribed above. We believe that the released passenger domain ismainly responsible for cleavage of C3 in fluid phase. However, C3cleavage by NalP expressed on bacterial surface may contribute toserum survival by reducing the generation of active C3 convertaseson themembrane. It will be of interest to determine if the inhibitionof complement activation is the result of increased degradation ofthe C3b-like molecule, decreased deposition of the C3b-like mol-ecule on bacterial surface, or decreased ability of the C3b-likemolecule to interact with factor B to form active C3/C5 convertases.Future studies will be needed to address the importance of NalP inthe pathogenesis of Nm infections and how the full-length and se-creted forms of NalP contribute to complement inhibition.In conclusion, Nm has evolved several redundant mechanisms

to evade the host innate responses at sites of colonization andduring systemic growth (7, 43). Several molecules, such as thecapsule, fHbp, and NspA, cooperate to confer serum resistance.Here, we have shown that NalP is an additional important mol-ecule that contributes to the arsenal of meningococcal defenses toevade the immune system.

Materials and MethodsBacterial Strains and DNA Manipulation. The strains, plasmids, and primers usedin this study are listed in Tables S2 and S3. Detailed descriptions of bacterialstrains, growth conditions, and methods for DNA manipulation and con-struction of Nm recombinant strains are provided in SI Materials and Methods.

Survival Experiments in Human Whole Blood and Human Serum. The experi-ments of survival in human blood and human serum were performed aspreviously described (14). Bacteria were grown until midexponential phaseand then added to human whole blood or human serum at 40% in D-PBS.Cultures were incubated for 2 h; at various time points, an aliquot of thesample was removed, and the number of viable bacteria was determined by

plating serial dilutions. Detailed methods and additional details are reportedin SI Materials and Methods.

Complement Cleavage Assay. Nm MC58 WT, NalP deletion mutant, andcomplementing strains were grown in Gonococcal (GC) Medium Base untilmidlog phase. Then, bacteria were centrifuged, the pellet was incubatedwithsingle complement factors (human factors C1q, C2, C3, and C4 and rabbit andmouse C3) for 16 h at 37 °C while shaking. Complement factor cleavage foreach factor was detected by Western blot analysis. The same experiment wasperformed using 50× concentrated bacterial supernatant. The specificity ofhuman C3 cleavage by NalP expressed on Nm bacterial surface was de-termined using the serine protease inhibitor PMSF. The activity of purifiedrabbit and mouse C3 was tested by checking rabbit or mouse C3 depositionon yeast glucan particles in presence of human C3-depleted serum (Fig. S6).Detailed methods are reported in SI Materials and Methods.

Analysis of C3 Fragments. After incubation of C3 with concentrated bacterialsupernatant from Nm NalP complementing strain, C3 fragments (C3a- andC3b-like molecules) were immunoprecipitated and then analyzed byWesternblotting. The C3b-like molecule was purified and subjected to N-terminalsequencing (Alphalyse). In addition, the C3b-like molecule was tested fordegradation by human host factors by adding fH and fI or 2% (vol/vol) humanC3-depleted serum. C3b-like molecule degradation was analyzed by Westernblotting. Additional details are reported in SI Materials and Methods.

C3b Deposition on Nm Bacterial Surface. Nm WT, NalP deletion mutant, andcomplementing strains were grown in GC medium until midlog phase. Then,bacteria were resuspended in 50% human serum diluted in HBSS++ and in-cubated at 37 °C for 60 min while shaking. C3b deposition was detected byFACS analysis. Additional details are reported in SI Materials and Methods.

Fig. 4. Schematic representation of the contribution of NalP in Nm comple-ment evasion. For NalP to function as a complement inhibitor, NalP full-lengthprotein on bacterial surface as well as the passenger domain containing thecatalytic site and secreted in the extracellular environment must both cleavehuman C3. Inhibition of complement activation by NalP occurs by cleavage ofC3, which generates a functionally inactive C3a-like molecule, whereas thecorresponding C3b-like fragment is inactivated rapidly by fH and fI. Thus, theactivity of NalP leads to the inactivation of C3 in human blood at surface leveland distal to the bacterial surface, resulting in the inhibition of C3b depositionand therefore, serum resistance.

Del Tordello et al. PNAS | January 7, 2014 | vol. 111 | no. 1 | 431

MICRO

BIOLO

GY

Dow

nloa

ded

by g

uest

on

May

23,

202

0

ACKNOWLEDGMENTS. We thank Alessandro Muzzi for the bioinformaticsanalysis of the nalP gene and Dr. Rosane DeOliveira for technical assistance.We thank Mariagrazia Pizza and Vega Masignani for useful discussion andGiorgio Corsi for artwork. I.V. is the recipient of a Novartis Fellowship from

the PhD Program in Functional Biology of Molecular and Cellular Systems ofthe University of Bologna. S.R. is supported by National Institute of Allergy andInfectious Diseases, National Institutes of Health Grants AI054544, AI084048,and AI32725.

1. Stephens DS, Greenwood B, Brandtzaeg P (2007) Epidemic meningitis, meningo-coccaemia, and Neisseria meningitidis. Lancet 369(9580):2196–2210.

2. Virji M (2009) Pathogenic neisseriae: Surface modulation, pathogenesis and infectioncontrol. Nat Rev Microbiol 7(4):274–286.

3. Figueroa JE, Densen P (1991) Infectious diseases associated with complement defi-ciencies. Clin Microbiol Rev 4(3):359–395.

4. Ram S, Lewis LA, Rice PA (2010) Infections of people with complement deficienciesand patients who have undergone splenectomy. Clin Microbiol Rev 23(4):740–780.

5. Jongerius I, Ram S, Rooijakkers S (2009) Bacterial complement escape. Adv Exp MedBiol 666:32–48.

6. Rambach G, Würzner R, Speth C (2008) Complement: An efficient sword of innateimmunity. Contrib Microbiol 15:78–100.

7. Serruto D, Rappuoli R, Scarselli M, Gros P, van Strijp JA (2010) Molecular mechanismsof complement evasion: Learning from staphylococci and meningococci. Nat RevMicrobiol 8(6):393–399.

8. Rooijakkers SH, van Strijp JA (2007) Bacterial complement evasion. Mol Immunol44(1-3):23–32.

9. Lambris JD, Ricklin D, Geisbrecht BV (2008) Complement evasion by human patho-gens. Nat Rev Microbiol 6(2):132–142.

10. Madico G, et al. (2006) The meningococcal vaccine candidate GNA1870 binds thecomplement regulatory protein factor H and enhances serum resistance. J Immunol177(1):501–510.

11. Seib KL, et al. (2009) Factor H-binding protein is important for meningococcal survivalin human whole blood and serum and in the presence of the antimicrobial peptideLL-37. Infect Immun 77(1):292–299.

12. Lewis LA, Carter M, Ram S (2012) The relative roles of factor H binding protein,neisserial surface protein A, and lipooligosaccharide sialylation in regulation of thealternative pathway of complement on meningococci. J Immunol 188(10):5063–5072.

13. Lewis LA, et al. (2010) The meningococcal vaccine candidate neisserial surface proteinA (NspA) binds to factor H and enhances meningococcal resistance to complement.PLoS Pathog 6(7):e1001027.

14. Echenique-Rivera H, et al. (2011) Transcriptome analysis of Neisseria meningitidis inhuman whole blood and mutagenesis studies identify virulence factors involved inblood survival. PLoS Pathog 7(5):e1002027.

15. Turner DP, Wooldridge KG, Ala’Aldeen DA (2002) Autotransported serine protease Aof Neisseria meningitidis: An immunogenic, surface-exposed outer membrane, andsecreted protein. Infect Immun 70(8):4447–4461.

16. van Ulsen P, et al. (2003) A Neisserial autotransporter NalP modulating the processingof other autotransporters. Mol Microbiol 50(3):1017–1030.

17. Saunders NJ, et al. (2000) Repeat-associated phase variable genes in the completegenome sequence of Neisseria meningitidis strain MC58. Mol Microbiol 37(1):207–215.

18. Roussel-Jazédé V, Grijpstra J, van Dam V, Tommassen J, van Ulsen P (2013) Lipidationof the autotransporter NalP of Neisseria meningitidis is required for its function in therelease of cell-surface-exposed proteins. Microbiology 159(Pt 2):286–295.

19. van Ulsen P, et al. (2006) A novel phase-variable autotransporter serine protease,AusI, of Neisseria meningitidis. Microbes Infect 8(8):2088–2097.

20. Serruto D, et al. (2010) Neisseria meningitidis GNA2132, a heparin-binding proteinthat induces protective immunity in humans. Proc Natl Acad Sci USA 107(8):3770–3775.

21. Roussel-Jazédé V, Jongerius I, Bos MP, Tommassen J, van Ulsen P (2010) NalP-mediated proteolytic release of lactoferrin-binding protein B from the meningo-coccal cell surface. Infect Immun 78(7):3083–3089.

22. Arenas J, Nijland R, Rodriguez FJ, Bosma TN, Tommassen J (2013) Involvement ofthree meningococcal surface-exposed proteins, the heparin-binding protein NhbA,

the α-peptide of IgA protease and the autotransporter protease NalP, in initiation ofbiofilm formation. Mol Microbiol 87(2):254–268.

23. Boulton IC, et al. (1999) Purified meningococcal transferrin-binding protein B inter-acts with a secondary, strain-specific, binding site in the N-terminal lobe of humantransferrin. Biochem J 339(Pt 1):143–149.

24. Granoff DM, Welsch JA, Ram S (2009) Binding of complement factor H (fH) to Neis-seria meningitidis is specific for human fH and inhibits complement activation by ratand rabbit sera. Infect Immun 77(2):764–769.

25. Terao Y, et al. (2008) Group A streptococcal cysteine protease degrades C3 (C3b) andcontributes to evasion of innate immunity. J Biol Chem 283(10):6253–6260.

26. Wexler DE, Chenoweth DE, Cleary PP (1985) Mechanism of action of the group Astreptococcal C5a inactivator. Proc Natl Acad Sci USA 82(23):8144–8148.

27. Park SY, et al. (2008) Immune evasion of Enterococcus faecalis by an extracellulargelatinase that cleaves C3 and iC3b. J Immunol 181(9):6328–6336.

28. Laarman AJ, et al. (2011) Staphylococcus aureus metalloprotease aureolysin cleavescomplement C3 to mediate immune evasion. J Immunol 186(11):6445–6453.

29. Orth D, et al. (2010) EspP, a serine protease of enterohemorrhagic Escherichia coli,impairs complement activation by cleaving complement factors C3/C3b and C5. InfectImmun 78(10):4294–4301.

30. Ramu P, et al. (2007) The surface protease PgtE of Salmonella enterica affects com-plement activity by proteolytically cleaving C3b, C4b and C5. FEBS Lett 581(9):1716–1720.

31. Gray-Owen SD, Schryvers AB (1993) The interaction of primate transferrins with re-ceptors on bacteria pathogenic to humans. Microb Pathog 14(5):389–398.

32. Sadarangani M, Pollard AJ, Gray-Owen SD (2011) Opa proteins and CEACAMs:Pathways of immune engagement for pathogenic Neisseria. FEMS Microbiol Rev35(3):498–514.

33. Sim RB, Twose TM, Paterson DS, Sim E (1981) The covalent-binding reaction ofcomplement component C3. Biochem J 193(1):115–127.

34. Daffern PJ, Pfeifer PH, Ember JA, Hugli TE (1995) C3a is a chemotaxin for humaneosinophils but not for neutrophils. I. C3a stimulation of neutrophils is secondary toeosinophil activation. J Exp Med 181(6):2119–2127.

35. Elsner J, et al. (1994) C3a activates reactive oxygen radical species production andintracellular calcium transients in human eosinophils. Eur J Immunol 24(3):518–522.

36. Elsner J, Oppermann M, Czech W, Kapp A (1994) C3a activates the respiratory burst inhuman polymorphonuclear neutrophilic leukocytes via pertussis toxin-sensitiveG-proteins. Blood 83(11):3324–3331.

37. Klos A, et al. (2009) The role of the anaphylatoxins in health and disease. MolImmunol 46(14):2753–2766.

38. Bokisch VA, Müller-Eberhard HJ (1970) Anaphylatoxin inactivator of human plasma:Its isolation and characterization as a carboxypeptidase. J Clin Invest 49(12):2427–2436.

39. Matthews KW, et al. (2004) Expression of the third complement component (C3) andcarboxypeptidase N small subunit (CPN1) during mouse embryonic development. DevComp Immunol 28(6):647–655.

40. Sayah S, et al. (2003) Two different transduction pathways are activated by C3a andC5a anaphylatoxins on astrocytes. Brain Res Mol Brain Res 112(1–2):53–60.

41. Oldfield NJ, et al. (2013) Prevalence and phase variable expression status of twoautotransporters, NalP and MspA, in carriage and disease isolates of Neisseria men-ingitidis. PLoS One 8(7):e69746.

42. Jain S, et al. (2006) Polar localization of the autotransporter family of large bacterialvirulence proteins. J Bacteriol 188(13):4841–4850.

43. Lo H, Tang CM, Exley RM (2009) Mechanisms of avoidance of host immunity byNeisseria meningitidis and its effect on vaccine development. Lancet Infect Dis 9(7):418–427.

432 | www.pnas.org/cgi/doi/10.1073/pnas.1321556111 Del Tordello et al.

Dow

nloa

ded

by g

uest

on

May

23,

202

0